[UPDATED]First look at the FuryX2

[Doobeedoo]

Seems odd for dual GPU to have cooling like that. But oh well, it might be. If it can be cooled well without lower clock than great.

[zMeul]

47 minutes ago, Exty said:

its a fury nano x2 atleast that is what amd twitter said before removing the pictures

the Nano has the same GPU as the Fury X, it's a Fiji XT

[Rob27shred]

Meh, dual GPU cards are & have always been a money grab gimmick IMO. It's no different than having 2 of the single GPU cards in SLI or CFX, in fact it might actually perform worse than that depending on the cooling setup. Until someone comes out with a dual GPU card that can use both GPUs without the drawbacks SLI or CFX I'll pass. 

[EliminatingAngels]

1 hour ago, Thiback said:

What's your source ?

I am quite dubious to this being the FuryX2. When dual gpu cards opt for a blower style cooler, they have their single fan at the center and not at the end, so as to blow air from both ends to the cards and thus keep both GPU evenly cooled.
The card on the photo is a blower style card, with the fan being at the end. If this was the Fury X2, this would mean that the chip closer to the end with the video outputs would run hotter than the one closer to the fan. 

I'd say this is a single GPU card, and I would speculate that this would be a Polaris GPU.

Its possible that the card is actually water cooled. If thats the case, having the blower fan at the end would be reasonable as it would pretty much only be cooling the vrm this way, it can cool and exhaust out the back. 

[Prysin]

 

 

[Watermelon Guy]

2 hours ago, Doobeedoo said:

Seems odd for dual GPU to have cooling like that. But oh well, it might be. If it can be cooled well without lower clock than great.

Its actually not that odd. Older GPUs often had this design.  like the 5970, 4870, gtx 290.

[Thiback]

31 minutes ago, Watermelon Guy said:

Its actually not that odd. Older GPUs often had this design.  like the 5970, 4870, gtx 290.

Seems I was wrong then. 

[Watermelon Guy]

15 minutes ago, Thiback said:

Seems I was wrong then. 

You were kind of right, gpus today use too much power to be cooled like that.

[Colonel_Gerdauf]

I really do want to see how they manage the power draw. The last time this issue was raised (with the 295X2), certain people were defending the product with silly anecdotes.

 

A specification violation is a specification violation, end of story. Whether or not the cards run well and safe, stock or overclocked, in specific case scenarios is a moot point and a useless argument.

[Prysin]

13 minutes ago, Colonel_Gerdauf said:

I really do want to see how they manage the power draw. The last time this issue was raised (with the 295X2), certain people were defending the product with silly anecdotes.

 

A specification violation is a specification violation, end of story. Whether or not the cards run well and safe, stock or overclocked, in specific case scenarios is a moot point and a useless argument.

said specification was dropped in March 2007. The "max 20 amp per rail" limit is no longer part of the official ATX spec.

Whilst the 8pin is "rated" at 150w, its maximum capacity is 296w.

 

When a bunch of college students and kids get their panties in a twist about "laws" governing things they do not fully understand. Well, i find it amusing to disprove you.

 

 

[Colonel_Gerdauf]

5 minutes ago, Prysin said:

said specification was dropped in March 2007. The "max 20 amp per rail" limit is no longer part of the official ATX spec.

Whilst the 8pin is "rated" at 150w, its maximum capacity is 296w.

 

When a bunch of college students and kids get their panties in a twist about "laws" governing things they do not fully understand. Well, i find it amusing to disprove you.

So... what does that have to do with the crux of the matter, exactly?

[Citadelen]

4 hours ago, zMeul said:

goes to their image and credibility

they're not confident enough to put an AMD CPU in that build

Or they're willing to accept that their CPU's aren't that great, which not many companies are willing to do.

[zMeul]

5 minutes ago, Citadelen said:

Or they're willing to accept that their CPU's aren't that great, which not many companies are willing to do.

then, they shouldn't complain and make video on how BAPCo Sysmark is unfair to their shit

[Prysin]

1 minute ago, Colonel_Gerdauf said:

So... what does that have to do with the crux of the matter, exactly?

you cannot violate a specification that is no longer officially valid. The end.

Plugs are rated by their manufacturer, and actually not specifically by the ATX spec, nor the MOLEX spec (6 and 8 pin falls under the MOLEX specs). Unfortunatly, people do not seem to understand that a "rating" is not a absolute. It is a guarantee of performance.

 

No matter how shit or low quality a 8pin is, it MUST meet the 150w requirement.

 

However, PSUs like Corsair AX and RM series, EVGA Gold, Platinum and Titanium PSUs and higher end Seasonic PSUs all come with thicker wires and sturdier connectors. you feel this when you try to insert them. Some connectors just slide in and out effortlessly. Higher end cables are minutely thicker and needs more force to be inserted or unplugged.

 

Now if AMD or Intel or Nvidia wants to sell a product with two 8 pins that consume 450w, then they are free to do so aslong as they do not violate the laws of electromechanics. Because THOSE are the laws that decides whether it will work, or "shits on fire yo".....

 

Oh and the OFFICIAL AMD spec of the R9 295x2 was 450w... but the subvendors like XFX/Sapphire/ASUS/Gigabyte/Powercolor ETC rated it higher.... ALL R9 295x2s are built by Sapphire btw. As is every other reference design card. So all Fury X and Fury Nano is Sapphire aswell).

 

 

[Citadelen]

Just now, zMeul said:

then, they shouldn't complain and make video on how BAPCo Sysmark is unfair to their shit

But it is an unfair representation of performance, why do people even bother arguing with you. Ehh, I guess I can reputation points making you look like an idiot.

[Colonel_Gerdauf]

19 minutes ago, Prysin said:

you cannot violate a specification that is no longer officially valid. The end.

Plugs are rated by their manufacturer, and actually not specifically by the ATX spec, nor the MOLEX spec (6 and 8 pin falls under the MOLEX specs). Unfortunatly, people do not seem to understand that a "rating" is not a absolute. It is a guarantee of performance.

 

No matter how shit or low quality a 8pin is, it MUST meet the 150w requirement.

 

However, PSUs like Corsair AX and RM series, EVGA Gold, Platinum and Titanium PSUs and higher end Seasonic PSUs all come with thicker wires and sturdier connectors. you feel this when you try to insert them. Some connectors just slide in and out effortlessly. Higher end cables are minutely thicker and needs more force to be inserted or unplugged.

 

Now if AMD or Intel or Nvidia wants to sell a product with two 8 pins that consume 450w, then they are free to do so aslong as they do not violate the laws of electromechanics. Because THOSE are the laws that decides whether it will work, or "shits on fire yo".....

 

Oh and the OFFICIAL AMD spec of the R9 295x2 was 450w... but the subvendors like XFX/Sapphire/ASUS/Gigabyte/Powercolor ETC rated it higher.... ALL R9 295x2s are built by Sapphire btw. As is every other reference design card. So all Fury X and Fury Nano is Sapphire aswell).

 

One, I need some empirical evidence for the fact that the specification is "no longer valid". To repeat, the fact that it is physically safe means absolutely nothing in the grand scheme of things. By this logic, AC Unity has a perfectly acceptable volume of draw calls, despite the breach of the "soft" limits. As you said yourself, this is largely dependent on the quality of the PSU that you use, but the specification takes everything into account when it comes to the power source and converter.

 

And two, you were giving completely different answers when this issue was originally brought up, so let me ask you; what gives?

[patrickjp93]

5 hours ago, TorqueS said:

Dunno why, but to me, it looks like a stock cooled 390 or 390X. I mean, I can't seem to see enough room for a radiator in that little case. Nor do I see any screws or mounts in that bottom right corner with the perforations.  

The stock cards are 11.61"(according to XFX's specs), which is roughly 30cm(29.48cm to be exact). That's about the length of the card in the picture.

 

EDIT:

NVM

 

 

 

So, the Fury X2 is NOT watercooled? The hell?

You can even see the blower in the provided picture. Maybe they'll allow custom PCBs and such...

[Godlygamer23]

21 hours ago, Master Disaster said:

That could be literally any card in there, I see nothing to indicate that's a dual fury card. I mean, it might be but you shouldn't always believe everything you read on the internet.

Looking at this photo, there's no context for how the whole thing looks. People say that it can't be a Fury X2 card because the fan is at the end. Maybe it's water-cooled and air cooled at the same time. The picture provided gives an awful view of the card and allows for no context, but a lot of speculation.

Edited February 29, 2016 by Godlygamer23
Capitalized "Fury".

[Citadelen]

Wait, how is the card facing towards us when the tower is standing up?

[Godlygamer23]

Just now, Citadelen said:

Wait, how is the card facing towards us when the tower is standing up?

PCIe extenders.

[Citadelen]

2 minutes ago, Godlygamer23 said:

PCIe extenders.

Ahh, I want one of those now, and thank you Godly, now I want to spend more money.

[Colonel_Gerdauf]

1 minute ago, Citadelen said:

Wait, how is the card facing towards us when the tower is standing up?

I highly doubt that this is a standard PC build. As Godly has said, they could very well be using extension cables.

[SubTract]

9 hours ago, Watermelon Guy said:

You were kind of right, gpus today use too much power to be cooled like that.

Well the 5970 was basically a 300W card.
So if the fury x2 (I really hope they don't call it the nano x2 since its defiantly not nano) falls into the same bracket its plausible to use a blower style cooler

I however feel like it is going to be water-cooled and just use the blower for VRM cooling.
Presuming they're using the same PCB Lisa Su showed off

They've got one hell of an extension hanging off the back of the card for a fan. 
2 water blocks would fit on there nicely.

Of course they wouldn't fit a fan between the 2 GPU's the way they have from the 6990. It could just mean that they didn't have another option.
The other thing looking at this image is the placement of the 2 8pin power connections. They would be blocked by an extension.

 

It could be that the actual Graphics card doesn't sit as far back it the case as the images lead us to believe. Perhaps a fake plastic shroud to deceive us all.

Of course they could have changed the board design significantly since then and it could all be codswallop.

[Prysin]

28 minutes ago, Colonel_Gerdauf said:

 

One, I need some empirical evidence for the fact that the specification is "no longer valid". To repeat, the fact that it is physically safe means absolutely nothing in the grand scheme of things. By this logic, AC Unity has a perfectly acceptable volume of draw calls, despite the breach of the "soft" limits. As you said yourself, this is largely dependent on the quality of the PSU that you use, but the specification takes everything into account when it comes to the power source and converter.

 

And two, you were giving completely different answers when this issue was originally brought up, so let me ask you; what gives?

 

 

Quote

ATX12V v2.3

Effective March 2007. Recommended efficiency was increased to 80% (with at least 70% required) and the 12 V minimum load requirement was lowered. Higher efficiency generally results in less power consumption (and less waste heat) and the 80% recommendation brings supplies in line with new Energy Star 4.0 mandates.^[20] The reduced load requirement allows compatibility with processors that draw very little power during startup.^[21] The absolute over-current limit of 240 VA per rail was removed, allowing 12 V lines to provide more than 20 A per rail.^{[citation needed]}

 

Quote

ATX12V v2.31

This revision became effective on February 2008. It added a maximum allowed ripple/noise specification of 400 millivolts to the PWR_ON and PWR_OK signals, requires that the DC power must hold for more than 1 millisecond after the PWR_OK signal drops, clarified country-specific input line harmonic content and electromagnetic compatibility requirements, added a section about Climate Savers, updated recommended power supply configuration charts, and updated the cross-regulation graphs.

ATX12V v2.32

This the unofficial name given to the later revisions of the v2.31 spec.^[22]

ATX12V v2.4

This is the current version of the ATX12V spec, published in April 2013. It is specified in Revision 1.31 of the 'Design Guide for Desktop Platform Form Factors',^[23] though it is also unofficially known as ATX12V v2.4.

Source:

https://en.wikipedia.org/wiki/ATX#Original_ATX

 

its a bit too late over here to start digging up the Molex specs. They're not at wikipedia, so you gotta google around to find em.

[Prysin]

31 minutes ago, Citadelen said:

Ahh, I want one of those now, and thank you Godly, now I want to spend more money.

you can get them on amazon for nearly nothing. They aint expensive.

Although, to find one that is ACTUALLY rated at Gen3 speeds takes some digging.... (it shouldnt matter, but meh, you know cheap knockoffs can always find a way to be shoddy)....